Cache is King
summary
Shows concrete wins using Ruby and Rails with MySQL Redis and Elasticsearch: replaces ActiveModelSerializers with Bulk Serialization and a cache class to drop 2,100 DB calls for 300 items down to seven and move RDS CPU from 100% to ~25%, swaps Redis lookups for a local hash to cut Redis hits and give ~65% faster jobs, uses the ActiveRecord connection object to avoid Redis for sharding, removes Resque throttling to stop Redis overload — real numbers and working fixes.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 4:51Kenna Security
- 6:35Bulk Serialization
- 11:46Redis Get Requests
- 14:43Local Cache vs Redis
- 17:47Solution
- 19:53Avoid
- 28:34Rescue Workers
